Abstract
This paper presents two control methodologies to obtain a robust performance of a robot manipulator. A dynamic model of the manipulator driven by three-phase induction motors is formulated. A torque control of one of the joints is presented. Torque control is very important, because you can determine the critical load that can be carried by the manipulator. Furthermore, using the inverse dynamics model it is possible to determine the positions and speeds of the manipulator joint. In this work, robust control techniques were implemented, such as Linear Quadratic Gaussian (LQG) and Linear Matrix Inequalities (LMI), and these two approaches are compared in performance. In addition, a Loop Transfer Recovery (LTR) procedure is used to achieve robustness to the uncertainties in the state estimation.